Most products get to market via a fast-track process; three experts debate whether the current system makes sense
By THOMAS M. BURTON FiDA highlight added
March 22, 2015 11:00 p.m. ET
Medical devices can save lives, but in recent years it has become clear how hazardous they are when they malfunction: metal hips that fail; bloodstream filters that fracture and prove lethal; defibrillator wires that break down.
The Food and Drug Administration works at striking the proper balance between insisting on safety but not standing in the way of innovative products that promote health and well-being. In doing so, it’s dealing with increasingly complex products of the software and digital age, including some aimed directly at consumers, such as medical apps.
The vast majority of medical devices still get to the U.S. market through an abbreviated pathway, called 510(k), in which companies need only show that a new device is substantially the equivalent of a device already legally marketed.
Joining our round-table discussion on medical-device safety are Scott Gottlieb, former FDA deputy commissioner and a resident fellow at the American Enterprise Institute; Rita Redberg, cardiology professor at UC San Francisco; and Bradley Merrill Thompson, device attorney in Washington, D.C. with Epstein, Becker & Green. Edited excerpts follow.
Does the system work?
WSJ: Does the FDA’s approach stymie innovation, or do patients need more protection?
DR. GOTTLIEB: FDA’s approach to regulating devices, in contrast to its requirements governing drugs, was designed by Congress to recognize that not all devices pose the same degree of risk. Therefore, the volume of data that the FDA requires should be closely matched to the risk of the product in question.
The problem today is the FDA has deviated from the original spirit of that idea. It’s trying to apply a much more uniform and druglike approach to its regulation of medical devices, increasing the hurdles that new products must clear. At the same time, the FDA is treating more low-risk devices like they were high risk.
One example where the FDA is trying to apply a more druglike approach to devices is in the growing use of “sham” surgeries. The FDA literally asks device makers to randomly assign patients to either get a new medical product, or undergo a fake surgery [as a placebo] to simulate the same procedure. In a recent trial, involving a device for the treatment of high blood pressure, patients were asked to undergo an angiogram—a procedure with risks—without knowing if they’d get the real treatment.
The FDA began asking some medical-device makers to use dummy surgical procedures a few years ago, to make its review of medical devices more like its review of drugs. Only in the case of devices, having a placebo group often can mean obligating a patient to a fake surgery. In that setting, consumers are being asked to take a risk without any prospect of a benefit. That seems unethical, and unnecessary.
The FDA also is showing a desire to classify more things that merely inform consumers as medical devices. One example is seen in the FDA’s approach to consumer apps, like those found on an iPhone. A lot of these apps are really low risk.
MR. THOMPSON: While there are certainly some incidents of the FDA regulating devices too stringently, they are by and large the exception. You mention placebo-controlled device trials. In the example you cite, however, the FDA did not order the company to use a placebo control. That was the company’s idea. The fact that the FDA permits the use of placebo controls is hardly evidence that the FDA is overregulating.
You suggest the FDA overregulates mobile apps. In the last couple of years, it has sought to reduce, not enlarge, its regulatory oversight. For example, it has deregulated the hardware and software that takes data from medical devices like a blood glucose meter and transmits it for storage and display. That was a big step forward.
Instead of sweeping, fundamental reform to either clamp down or loosen up FDA oversight, I favor continuous improvement guided by continuous review of agency performance data as well as marketplace and scientific trends.
DR. REDBERG: As a cardiologist taking care of patients every week, I see too many whose lives have been harmed and who have suffered greatly from untested or inappropriate devices. And that doesn’t begin to address the billions of dollars these devices add to health-care bills.
Only a few percent of all devices on the U.S. market go though the only pathway that requires any clinical data. And even that most-stringent pathway doesn’t require two randomized clinical trials. In fact it doesn’t even require one; this most rigorous evidence standard can and frequently is met by non-randomized trials without controls, or by case-series or even expert opinion.
These devices cannot be removed without significant risk of serious morbidity or death, as has been learned by the many women who received pelvic mesh, or had their uteri removed with the aid of morcellators, for just two of many possible examples. Yet another example of a high-risk implanted device entering the market without clinical studies is the inferior vena cava filter, which for some models, studies have shown that 1 in 4 fracture after implantation in the major blood vessel entering the heart and are risky to remove. And after metal-on-metal hips, touted as a great innovation, entered the U.S. market without clinical studies, many Americans learned the hard way that they have a fourfold increased revision rate compared with other artificial hips and leak metal ions into the bloodstream.
WSJ: Brad and Scott, isn’t it clear that too often, devices get on the market with minimal evidence of safety?
DR. GOTTLIEB: My concern is that a lot of the fundamental questions about device safety turn on issues of biomechanical performance. For example, some common questions are how resilient an implanted valve will be to the shearing effects of blood flow, a joint to constant stresses, or the durability of a pacemaker’s circuits.
The FDA is trying, in too many cases, to answer these questions by requiring longer and larger trials in people and large animals like pigs and sheep.
The issue at the heart of the tragedy with [uterine] morcellators [in hysterectomies] is a different one. Here there was a much more fundamental misunderstanding about the biology of a disease state. The FDA didn’t fully appreciate the risk of certain kinds of cancer, and how frequently they occurred.
MR. THOMPSON: The truth, I think, lies somewhere in between the extremes that [Dr. Gottlieb] and [Dr. Redberg] have staked out. I agree with [Dr. Redberg] that sometimes unsafe or ineffective products reach the market. Morcellators are indeed a recent example. That’s regrettable. However, that fact, by itself, doesn’t mean we should fundamentally change the system.
Out of all of the thousands of medical devices that the FDA clears each year, how many turn out to be unsafe or ineffective? Not many. But further, and equally importantly, how effective are alternative regulatory systems in approving important new medical devices in a timely way? Changing the system in a way that delays or even forecloses the sale of safe and effective breakthrough medical devices hurts patients, too.
WSJ: In one of the worst failures, a Medtronic defibrillator wire, the company relied on stress testing, not a clinical trial; 268,000 patients were exposed to a wire that led to multiple deaths.
MR. THOMPSON: That problem was found eight years ago. The fact you need to go back that far speaks to the overall robustness of the FDA process.
DR. GOTTLIEB: The failure of any defibrillator lead is tragic, but to portray the malfunction of Medtronic’s lead as an example of a device failure that could have been avoided through a bigger clinical trial gives short shrift to the basic science and engineering challenges that permeate this type of problem, and how one tries to prevent it from occurring.
Keep in mind that 99.8% of all medical devices have no serious adverse events associated with them. The U.S. is the gold standard when it comes to device safety. There are always opportunities to enhance the process. For example, the FDA is implementing a unique device identifier [UDI] system to better track device performance [after approval].
DR. REDBERG: Only a few percent of all medical devices enter the market via a pathway that requires at least some clinical data.
Simply put, for the vast majority of medical devices, there is NO requirement to demonstrate safety and effectiveness.
We must have national, publicly available registries of devices and procedures, so that we can track in real-time the outcomes, and then truly know the benefits and the risks.
Mr. Burton is a reporter in the Washington bureau of The Wall Street Journal. He can be reached at email@example.com.
"The Wall Street Journal and Thomas M. Burton have my gratitude for tackling this issue, but my dismay for having a ‘roundtable’ discussion that does not include the patient perspective.You do not need to go back eight years to find implanted medical devices failing. Nearly 70,000 women are in federal court in an MDL lawsuit for failed pelvic mesh sold by more than a half dozen U.S. medical device makers. 14,000 women have met on Facebook Essure Problems to provide mutual support and begin the legal and legislative process of removing this dangerous Bayer ‘permanent’ birth control product from the market. The company is profiting while they suppress information and harm more young women and their families.As a trained FDA Patient Representative (9/2010) I learned what little regard the clearance process had for PREVENTABLE harm to patients. In the ‘Informed Consent’ there is no access to non-proprietary patient outcome research on the device and no discussion of the impact of “Riegel v Medtronic” and pre-emption on the civil rights of the patient should he/she be harmed. There is no product warranty and hospitals and doctors have a back-door way of dealing with the complex medical issues of a FAILED device: the patient is referred for psychiatric help, a pain clinic or the doctor/patient relationship is severed by the doctor. Harmed patients are buried alive by the perpetrators-in plain sight.” This is legal, thanks to aggressive lobbyists for the medical device (petroleum, chemical, metals) industry.